Blogs by Category:
Metadata-driven ETL/ELT, CDC, and workflow automation that keeps pipelines predictable and repeatable.
October 5, 2026
A failed batch leaves skip flags on every package that finished. If nobody reruns it, the next scheduled run skips them and logs Success. Here is how to find those flags and set an expiry window.
September 28, 2026
A working definition with the artifacts to back it: the same batch generated for SSIS, Azure Data Factory and Databricks, the run log every step writes, and a three-question checklist for judging tools.
September 4, 2026
A failed load leaves one recovery flag in the BimlCatalog you own. The next run reads that flag and does the strongest thing its runtime supports: a generated rollback container on SSIS, a clean re-execute on ADF.
August 28, 2026
How BimlFlex prices every load by the activities it needs, packs generated sub-pipelines against a configurable budget, chains the overflow, and fails the build with a clear message instead of failing your deployment.
August 12, 2026
After each successful batch, BimlCatalog reads the timing of the last two successful runs and records which package finished right before each one started. See the mechanism, its limits, and the query that reads the result back.
July 29, 2026
Each batch in your model becomes a generated job file where solve-order groups run as chained sub-jobs and per-object thread numbers decide what runs in parallel. The dependency logic lives in YAML you can read, diff, and version.